"Howard J. Rogers" <hjr_at_dizwell.com> wrote in message news:<40b6d4e4$0$3032$afc38c87_at_news.optusnet.com.au>...
> "yls177" <yls177_at_hotmail.com> wrote in message
> news:c06e4d68.0405272114.5bc35034_at_posting.google.com...
> > size of one extent is 2097,080 kb
> > 3 extents will have 6291240
> >
> > never used blocks = 112510,block size is 8kb, therefore in kb is
> > 4817024
> >
> > hence, if add 3 more extents, the tablespace will overflow.
> >
> >
> > however, the above analysis is based that there are no big enough free
> > contiguous space in used blocks for the size of one extent, if there
> > is, then this could change the situation right?
>
> Well, it could if you don't do your space management properly. Since at
> least version 7, Oracle has been trying to say to DBAs, over and over again:
> please only use one extent size per tablespace. In 8.0 they invented the
> MINIMUM EXTENT clause to make it happen. In 8i, they invented the
> uniform-sized locally managed tablespace to guarantee to make it happen.
>
> If you had followed that advice, then the question of there being 'no big
> enough free contiguous space' would never arise. All chunks of free space
> would be extent-sized, and therefore the problem would disappear.
>
> It's difficuly to comment further without knowing your Oracle version,
> though.
>
> > and hence
> > i get the value for " never used blocks" which guarantee that they are
> > contiguous
>
> I can't say whether that guarantees anything, because I don't know where you
> got that number from. Care to elaborate?
>
> Regards
> HJR
i get these figures from sap db02, tcodes... therefore, i am not sure
of the underlying sql statements...
Received on Sun Jun 06 2004 - 07:16:21 CDT
